The antiandrogenic and gestagenic steroid cyproterone acetate (CPA) has been widely used in human therapy. There is currently a debate about the safety of CPA, since it proved to be genotoxic in rat liver and human hepatocytes [I. Neumann et al., Carcinogenesis (Lond.), 13: 373–378, 1992; J. Topinka et al., Carcinogenesis (Lond.), 14: 423–427, 1993; L. R. Schwarz et al., Biological Reactive Intermediates: V. Basic Mechanistic Research in Toxicology and Human Risk Assessment, pp. 243–251, 1996; A. Martelli et al., Carcinogenesis (Lond.), 16: 1265–1269, 1995].
Little is known about the metabolic pathways of activation of CPA to genotoxic metabolites. Using rat hepatocytes and subcellular fractions of female rat liver, we have examined whether sulfoconjugation plays an essential role in the activation of CPA to DNA-binding metabolites which are detectable with 32P-postlabeling. Incubation of hepatocyte cultures with 30 µm CPA for 6 h caused the formation of several DNA adducts; the total adduct level amounted to about 12,400 adducts/109 nucleotides. When the cells were incubated in sulfate-free medium to prevent the synthesis of the cosubstrate of sulfonation, 3′-phosphoadenosine-5′-phosphosulfate (PAPS), formation of all CPA-DNA adducts was greatly reduced, amounting to only 5% of that determined in the presence of sulfate (810 µM). Activation of CPA is likely to be catalyzed by hydroxysteroid sulfotransferase(s), because the specific substrate dehydroepiandrosterone almost completely inhibited DNA-binding of CPA. Our assumption that sulfonation plays a decisive role in the bioactivation of CPA is further supported by the results obtained with an in vitro system consisting of calf thymus DNA, various subcellular liver fractions, and the cofactor PAPS, NADPH, or NADH. Significant DNA binding only occurred when cytosol and both PAPS and the reduced pyridine nucleotides were present. The DNA adduct spot obtained was chromatographically identical to the adduct spot A detected in isolated liver cells, suggesting that the CPA-DNA adduct formed in vivo and in vitro is identical. Cytosol is known to contain not only sulfotransferases but also reductases. Thus, the requirement for NADPH or NADH suggests that in addition to sulfotransferase(s), reductases are involved in the activation of CPA.
We propose that bioactivation of CPA involves reduction of the keto group at C-3 followed by sulfonation of the hydroxysteroid. The resulting sulfoconjugate is most likely unstable and supposed to generate a reactive carbonium ion.